1. Field of the Invention
The present invention relates to a communication field, and particularly to a method for re-balancing a load, a method for migrating a load, a device and a system using the same.
2. Description of the Prior Art
A core network (CN) pool area refers to multiple wireless areas served by a group of same type CN nodes such as SGSN (Servicing GPRS Support Node) or MSC (Mobile Switching Center) or MME (Mobile Management Entity) simultaneously.
When a group of CN nodes serve one pool area, as shown in FIG. 1, each RAN (Radio Access Network) node (such as RNC (Radio Network Controller)/BSC (Base Station Controller)) in the pool area and CN nodes (such as MSC (Mobile Switching Center)) of the same type wireless system are all interconnected and multiple CN nodes share access requests from UEs (User Equipment) in the pool area. When any UE initially enters the pool area, the RAN node has NNSF (NAS, Node Selection Function) of NAS (Non access Stratum) and selects the CN node according to a load balancing principle and forwards the access request to corresponding CN node. In addition, the RAN node may set a WF (Weight Factor) according to capacity differences of CN nodes so that the CN node may be selected according to the difference between WFs.
When a UE moves in the pool area, the CN node serving the UE remains the same. After the UE registers in the CN, the CN node may allocate a temporary identifier for the UE. For the PS (Packet Switching) domain network of 2 G/3 G, the temporary identifier is P-TMSI (Packet-Temporary Mobile Subscriber Identity), while for the CS (Circuit Switching) domain network of 2 G/3 G, the temporary identifier is TMSI (Temporary Mobile Subscriber Identity). SGSN employs NRI (Network Resource Identifier) field in P-TMSI to identify itself, MSC employs NRI field in TMSI to identify itself, and MME (Mobility Management Entity) employs MMEI (Mobility Management Entity Identifier) field in GUTI (Globally Unique Temporary Identity) to identify itself. If the UE sends the access request in the same area once more after receiving the temporary identifier, the RAN node will select the CN node served last time according to the NRI or MMEI.
If a certain CN node serving in a pool area needs to migrate the UE it is serving to another CN node in the same pool area, it will use the load re-balancing technique.
For 2 G/3 G network, the CN node (SGSN or MSN) may set the NRI field to NULL NRI (Null Network Resource Identifier) or designate the NRI field to an identifier of another CN node and then feeds back the result to the UE so as to urge the UE to initiate a second access request as soon as possible. When the UE initiates the second access request, the RAN node will not forward the second access request to the CN node serving last time.
For LTE (Long Term Evolution) network, the CN node (MME) sets MMEI as an identifier of another CN node and feeds back the result to the UE so as to urge the UE to initiate a second access request as soon as possible. Or the MME may signal the RAN node that the load re-balancing is requested by the UE. Then the RAN node will notify the UE of the fact and urge the UE to initiate a second access request. When the UE initiates the second access request, the RAN node will not forward the second access request to the CN node serving last time.
The conventional technique has at least the following disadvantages: according to a conventional load re-balancing mechanism, the CN node may migrate the UE only according to a load status and for a specified UE with a special need, no specific process has been provided.